Relationship Between Endothelial Dependent Vasodilation and Size of Abdominal Aortic Aneurysms
Article Outline
Background
Flow-mediated dilation of the brachial artery (FMDB) indirectly reflects the action of nitric oxide liberated by the endothelium. In patients with abdominal aortic aneurysms (AAA), the changes in nitric oxide metabolism in close association with inflammation, appear to play a leading role in the aetiopathology of this disease, although it is still not clear. The objective was to study the correlation and behavior of FMDB relative to the aneurysm diameter (AD). To evaluate the relationship between C-reactive protein (CRP) and the FMDB in these patients. To study the correlation and behavior of FMDB relative to the aneurysm diameter (AD). To evaluate the relationship between C-reactive protein (CRP) and the FMDB in these patients.
Methods
The FMDB value and the CRP were determined in a total of 30 patients with an AAA ≥30 mm, confirmed by computed tomography. The cardiovascular and treatment history was recorded, together with the lipid and renal profile and leucocyte count.
Results
The median AD in the sample was 43 mm (25 percentile: 37 mm; 75 percentile: 60 mm). The primary variables of the study, FMDB and CRP, were the only ones that differed statistically when we stratified the sample according to AD quartiles (p < 0.001). There was a negative correlation between the FMDB and the AD (R = −0.78 [p < 0.001]) and a positive one for CRP (0.74 [p < 0.001]). The CRP/FMDB gave an R value of −0.56 (p = 0.001).
Conclusion
Endothelium dependent vasodilation has a linear and negative correlation with the AD. The positive correlation between the FMDB and CRP supports the hypothesis that inflammation and endothelial dysfunction are processes associated with the physiopathology of AAA and vary with their growth.
Introduction
Despite being widely studied, the aetiopathology of abdominal aortic aneurysms (AAA) is not completely known. Many studies have shown that their origin is due to many factors, for example, the destruction of elastin and collagen fibres,1, 2 increase in metalloproteinases of the matrix,3, 4, 5 the existence of a chronic inflammatory process,6, 7, 8 and the increase in apoptosis of the smooth muscle cells.9, 10
Changes in nitric oxide (NO) are shown to have an increasingly important role in inflammation and the aetiopathology of AAA.11, 12, 13, 14, 15, 16 Furthermore, these changes are considered to be one the earliest events in arteriosclerotic disease.17, 18
The systematic nature of arteriosclerosis, with its signs at many levels, continues to be a dilemma as to what factors determine that an artery is dilated or has, on the other hand, a stenosis, when it is assumed that arteriosclerosis is the underlying process in both conditions.15, 17, 18, 19, 20
In our previous studies on peripheral arterial disease (PAD), although showing a change in the endothelial-dependent vasodilation determined by flow-mediated dilation of the brachial artery (FMDB), this did not correlate with the different clinical spectra of the disease, being independent of its severity, and reinforcing the idea of its early involvement in the development of PAD.21, 22
In this study, we consider whether the AAA classically called arteriosclerotic would show another pattern in connection with endothelial-dependent vasodilation other than that of PAD.
Objective
To study the correlation and behavior of FMDB relative to the size of the infrarenal abdominal aortic aneurysms, and to evaluate the relationship between the high-sensitivity C-reactive protein (CRP) and FMDB values.
Materials and Methods
A cross-sectional observational study that included 30 patients with an infrarenal AAA of 3 cm or more in diameter, diagnosed using computed axial tomography. The aneurysm diameter (AD) was obtained from the largest cross-section perpendicular to the longitudinal axis of the aorta using the three-dimensional measurement tools of the DICOM OsiriX image management software (© Antoine Rosset), version 2.6.
Inflammatory or saccular aneurysms, and those that were symptomatic, were excluded. Patients with acute or chronic inflammations, or were on steroid treatment, were also excluded.
Endothelial-dependent vasodilation was determined on all patients by measuring the FMDB with ultrasound, following the routine technique described in previous studies by our group.21 A single investigator who did not know the AD values performed all the FMDB determinations. The ankle-brachial index, the lipid and renal profile, and full blood count were determined, as well as the CRP levels using high sensitivity automated immunoassay (Roche Diagnostics), with a lower detection limit of 0.2 mg/L and a coefficient of variation of 4.2% at 4 mg/L and 6.3% at 1 mg/L.23 The main risk factors and treatments were recorded. The intra-observer coefficient of variation in the measurement of the FMDB was 2.76%.21
Patients with a baseline blood glucose greater than 120 g/dL or those who required glucose lowering treatment were considered as diabetics.24 Those patients with systolic values greater than 140 mm Hg and diastolic values greater than 90 mm Hg, or who were on hypertension treatment, were considered to be hypertensive. Hyperlipidaemia was established as a total plasma cholesterol value over 240 mg/dL, an LDL cholesterol over 160 mg/dL, and triglycerides greater than 200 mg/dL, or if the patient was taking lipid lowering drugs.25 A plasma creatinine level greater than 1.5 mg/dL was considered as chronic renal failure.26 The presence of PAD was assumed in those patients with absence of pulses and an ankle-brachial index of <0.9.
Statistical Analysis
The normality of the distribution of the variables was determined using the Kolmogorov—Smirnov test. The normal distribution variables were compared using the Student t test and analysis of variance (“post hoc” analysis: Tukey test), with the Mann Whitney U and Kruskal-Wallis tests being used for non-normal distributions. Multivariate analysis was performed on the variables that showed significant differences using linear multiple regression. The categorical variables were compared using the chi-squared test, with the exact Fisher test being applied if required. Spearman's Rho test was used for the correlation analysis. Receiver operating characteristic curves were used to determine the cut-off points. The variables are expressed as mean and 95% confidence interval, if there was a normal distribution, and as median and 25 and 75 percentiles, if there was not. The sample size was calculated by extracting a pilot sample before conducting the study, to obtain a power of 80% with a beta error of 0.5 for the expected differences. Based on the statistics obtained on our pilot sample, the sample required was 24. A p < 0.05 was considered statistically significant. All the calculations were performed using SPSS 11 statistics package.
Results
All the patients were male, with a mean age of 67.83 years (64.89-70.77). The median AD was 43 mm (37; 60). The mean value of the FMDB was 5.02% (3.92-6.12). The CRP median was 4.15 mg/L (3; 6.7).
The FMDB was 2.47% (1.01-3.93) for aneurysms greater than 55 mm and 6.29% (5.13-7.4) for those with a smaller AD. Table I shows the values of the variables between the groups stratified according to the AAA diameter quartiles. Only the FMDB and the CPR were significantly different between themselves. The multivariate analysis did not show independence between any of these variables in their relationship with the AD (p < 0.001). The Tukey test analysis showed that the difference between the FMDB values was not significant between the adjacent quartiles, but it was between the subsequent ones, as shown in Figure 1.
Table I. Variables according to aneurysm diameter quartiles
| Groups according aneurysm diameter quartiles | ≤37 mm n = 8 | 38-43 mm n = 7 | 44-59 mm n = 7 | ≥60 mm n = 8 | p |
|---|---|---|---|---|---|
| Age (yr) | 67.38 (59.5-75.25) | 66.71 (58.73-74.7) | 68.29 (61.17-75.4) | 68.88 (62.96-75.25) | 0.95 |
| BBD (mm) | 4.69 (4.14-5.24) | 4.84 (4.18-5.5) | 4.9 (4.43-5.38) | 4.75 (4.07-5.43) | 0.93 |
| FMDB (%) | 8.05 (5.81-10.29) | 6.07 (4.92-7.2) | 3.74 (2.69-4.8) | 2.19 (0.48-3.9) | <0.001 |
| CRP (mg/dL) | 3 (2.57; 3.25) | 4.1 (3.5; 4.96) | 4.2 (2.4; 4.2) | 9.6 (6.77; 12.47) | <0.001 |
| Leucocytes/mm3 | 7,912 (7,264; 8,550) | 8,176 (7,250; 11,300) | 7,900 (7,700; 8,400) | 7,945 (7,055; 9,337) | 0.94 |
| (n/%) | |||||
| Risk factors | |||||
 Hypertension | 5 (62.5) | 3 (42.9) | 4 (57.1) | 6 (75) | 0.64 |
| Diabetes mellitus | 4 (50) | 1 (14.3) | 1 (14.3) | 2 (25) | 0.34 |
| CAD | 2 (25) | 1 (14.3) | 1 (14.3) | 1 (12.5) | 0.9 |
| Smoker | 2 (25) | 4 (57.1) | 3 (42.9) | 4 (50) | 0.61 |
| CRF | 1 (12.5) | 0 (0) | 0 (0) | 1 (12.5) | 0.59 |
| COPD | 1 (12.5) | 2 (25) | 0 (0) | 1 (12.5) | 0.47 |
| Dyslipemia | 7 (87.5) | 2 (28.6) | 5 (71.4) | 5 (62.5) | 0.12 |
| CVA | 4 (50) | 1 (14.3) | 0 (0) | 2 (25) | 0.41 |
| PAD | 3 (37.5) | 4 (57.1) | 1 (14.3) | 2 (25) | 0.35 |
| Medical therapy | |||||
| ACE-I | 4 (50) | 3 (42.9) | 3 (42.9) | 5 (62.5) | 0.85 |
| CCB | 0 (0) | 0 (0) | 2 (25) | 2 (25) | 0.33 |
| Beta- blockers | 1 (12.5) | 0 (0) | 1 (12.5) | 3 (42.9) | 0.25 |
| Statins | 5 (62.5) | 1 (12.5) | 2 (28.6) | 2 (28.6) | 0.21 |
| Anti-platelet drugs | 6 (75) | 5 (62.5) | 3 (42.9) | 4 (57.1) | 0.5 |
| Anticoagulants | 0 (0) | 0 (0) | 1 (12.5) | 0 (0) | 0.33 |
Fig. 1
Box diagram of the flow-mediated dilation of the brachial artery (FMDB) values of the different aneurysm diameter (AD) quartiles. The arrows show those that had a p < 0.05 between them after the Tukey analysis.
The study of the correlations between the FMDB and the AD gave a correlation coefficient of r = −0.78 with a p < 0.0001 (Fig. 2). The CRP and the AD showed an r = 0.74 with p < 0.0001 (Fig. 3). There was also a correlation between the CRP and the FMDB with r = −0.56, p = 0.001 (Fig. 4).
Fig. 2
Scatter and linear diagram of the central tendencies of the FMDB and AD variables. Spearman Rho correlation coefficient R = −0.78 (p < 0.0001).
Fig. 3
Scatter and linear diagram of the central tendencies of the C-reactive protein (CRP) and AD variables. Spearman Rho correlation coefficient R = 0.71 (p < 0.0001).
Fig. 4
Scatter and linear diagram of the central tendencies of the FMDB and CRP variables. Spearman Rho correlation coefficient R = −0.56 (p < 0.0001).
In the receiver operating characteristic curve analysis, and establishing the presence of an aneurysm of 55 mm or more as a positive test, the area under the curve was 87% (p = 0.001). An FMDB value lower than 3.7% was associated with an aneurysm greater than or equal to 55 mm, with a sensitivity of 70% and a specificity of 80% in our sample (Fig. 5). An AD cut off of 55 mm was chosen as this is the current commonly accepted limit in which surgical repair offers benefits against the risk of rupture, and better illustrates the possible clinical usefulness of measuring the FMDB.
Fig. 5
Receiver operating characteristic curve to predict the FMDB of an abdominal aortic aneurysms of 55 mm or more. Area under the curve 87% (p = 0.001). Cut off point: 3.7% (Sensitivity 70%; Specificity 80%).
Discussion
The results of this study support the importance of NO disturbances in aneurysm disease. There is a clear difference in the FMDB values between patients with aneurysms with a size in the surgical range (>55 mm) and those of a smaller diameter. This persists on stratifying the sample into groups according to the AD quartiles. The Tukey analysis shows us that this difference not only exists between extreme diameter values, but also between alternate quartiles as shown in Figure 1. The correlation analysis added a strong association between the decrease in vasoreactivity and the increase in the size of the aneurysm. Only Knipp et al. have evaluated the existence of changes in the vasoreactivity of the brachial artery in relation to the size of the AAA. They analyzed this variable in 10 control patients, nine patients with PAD and 11 with AAA. This group did not find any correlation between the FMDB and AD, although the number of patients was small for this analysis, as they acknowledged in their discussion. They gave no details of the methodology or the results that supported this assertion.13
In a previous study by our group, the CRP positively correlated with the AD in a group of 67 patients with AAA.8 Furthermore, this condition is increasingly being established in the literature as a chronic inflammatory process.5, 6, 7 The current study confirms our previous findings, and also establishes a significant negative correlation between the CRP and FMDB values, which, as the studies on these indicate, reinforces the idea that inflammation and endothelial dysfunction are closely linked processes in the physiopathology of the aneurysm.5, 6, 16, 18 In fact, CRP is the only variable, along with the FMDB, that differs between the groups analyzed, even after the multivariate analysis. This gives the study external consistency.
In view of the finding of a decrease in systemic vasoreactivity, as shown by the decrease in FMDB, and the fact that aneurysm disease is often present in multiple locations, endothelial dysfunction seems to be a key factor in the development of the disease, leading to a vascular tree prone to dilation. This point should be demonstrated in further studies to confirm whether the decrease in vasoreactivity persists or not after resection of the aneurysm, although the association found between FMDB and AD in this study reinforces this idea.
It continues to be an enigma that PAD and AAA, sharing endothelial dysfunction and inflammation as pathogenic substrates, behave so differently. The importance appears to lie in that NO homeostasis is essential for remodeling the arterial wall by controlling the growth of smooth muscle cells. When the levels of NO increase above the homeostatic values they suffer from peroxidation and cause disturbances in the renewal of smooth muscle cells, leading to massive apoptosis and destruction of the structural fibres of the wall, with its subsequent degeneration.14, 16
It should be emphasized that the vasoreactivity disturbances are highly associated with the AD, whereas in PAD, although there is an overall decrease, the FMDB values are similar between patients with intermittent claudication and those affected by critical ischaemia, and independent of the ankle-brachial index values.21, 22
The present study has a limitation in that it is unable to discern, in any way, the pathogenic “significance” of the CRP value, which could be associated either to PAD or aneurysm disease. In any case, there were no significant differences between the groups analyzed according to the AD quartiles as regards the classic cardiovascular disease “group”: coronary disease, cerebrovascular disease, and PAD. Thus confirming, for the analysis of the CRP values found between these groups, the proportion of this CRP that we could assume contributes to classic cardiovascular disease itself. Even though, if we look carefully, we can observe both groups with a greater AD and in those who, in turn, we have found a higher CRP concentration have, contrary to expectations, a lower rate of PAD and ischaemic heart disease, diseases where an early and delayed association with CRP levels has been demonstrated. Thus, it is shown that the increased CRP levels found directly associated with the AD are a result of aneurysm disease and not of PAD or ischaemic heart disease. This different behavior that we can observe in arterial vasoreactivity and the inflammatory markers in aneurysm disease all but confirms that this disease is just another clinical variant of arteriosclerosis with a deviation in the pathogenic pathway that makes it different from PAD.
All this highlights the complexity of the mechanisms involved in the aetiopathology of arteriosclerosis, which after decades of studies still needs to be clarified, with discussions in favor of a multifactorial origin and a need for an increase in studies on the aetiology of the disease at a more basic molecular level.
References
- . Pathogenesis of abdominal aortic aneurysms: cellular and biochemical mechanisms. Eur J Vasc Endovasc Surg. 1996;12:391–400
- Pathogenesis of abdominal aortic aneurysms: a multidisciplinary research program supported by the National Heart, Lung and Blood Institute. J Vasc Surg. 2001;34:730–738
- Cellular localization of matrix metalloproteinases in the abdominal aortic aneurysm wall. J Vasc Surg. 1994;20:814–820
- . Expressions of matrix metalloproteinases in human abdominal aortic aneurysm tissue. Chin J Exp Surg. 1999;16:26–27
- . Inflammation, metalloproteinases, and increased proteolysis: an emerging pathophysiological paradigm in aortic aneurysm. Circulation. 1997;96:2115–2117
- . The role of inflammation in non-specific abdominal aortic aneurysm disease. Ann Vasc Surg. 1991;5:229–233
- . Effects of inflammatory infiltration on the formation of abdominal aortic aneurysm. Chin J Surg. 1999;37:177–179
- . Relación del tamaño del aneurisma de aorta abdominal asintomático y los niveles plasmáticos de proteína C reactiva. Angiología. 2007;59:111–120
- . Smooth muscle cell elastase, atherosclerosis, and abdominal aortic aneurysms. Ann Surg. 1992;216:327–330
- . Death of smooth muscle cells and expression of mediators of apoptosis by T lymphocytes in human abdominal aortic aneurysms. Circulation. 1999;99:96–104
- . Inhibition of inducible nitric oxide synthase limits nitric oxide production and experimental aneurysm expansion. J Vasc Surg. 2001;33:579–586
- . Nitric oxide in experimental aneurysm formation: early events and consequences of nitric oxide inhibition. Ann Vasc Surg. 2002;16:65–72
- Impaired vasoreactivity despite an increase in plasma nitrite in patients with abdominal aortic aneurysms. J Vasc Surg. 2002;35:363–367
- . Inducible nitric oxide synthase is present in human abdominal aortic aneurysm and promotes oxidative vascular injury. J Vasc Surg. 2003;38:360–367
- Relationship between iNOS expression and aortic cell proliferation and apoptosis in an elastase-induced model of aorta aneurysm and the effect of 1400 W administration. Surgery. 2005;137:447–456
- Role of nitric oxide and inducible nitric oxide synthase in human abdominal aortic aneurysms: a preliminary study. Chin Med J. 2006;119:312–318
- . Role of endothelial dysfunction in atherosclerosis. Circulation. 2004;109:III-27–III-32
- . Nitric oxide and arterial disease. J Vasc Surg. 2004;40:187–193
- Vascular reactivity in patients with peripheral vascular disease. Am J Cardiol. 1995;76:207–212
- . Abdominal aortic aneurysm: a general defect in the vasculature with focal manifestations in the abdominal aorta?. J Vasc Surg. 1997;26:247–254
- . Acı'n F. Relationship between non invasively measured endothelial function and peripheral arterial disease. Angiology. 2009;60:725–731
- . Relación de la dilatación de la arteria braquial mediada por flujo y el índice tobillo-brazo en pacientes con enfermedad arterial periférica. Angiología. 2007;59:55–61
- . Development of a new microparticle-enhanced turbidimetric assay for C-reactive protein with superior features in analytical sensitivity and dynamic range. J Clin Lab Anal. 1998;12:137–144
- Report of the expert committee o n the diagnosis and classification of diabetes mellitus. Diabetes Care. 2001;24:S5–S20
- National Heart, Lung, and Blood Institute. Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III, or ATP III), 2001. NIH Publication No. 01-3305. Available at http://www.nhlbi.nih.gov/guidelines/cholesterol/atglance.htm.
- . Chronic renal insufficiency. Part I: definition, clinical course stages, progression mechanisms, etiology, and diagnostic criteria. Nefrología. 2001;21:18–20
PII: S0890-5096(10)00155-X
doi:10.1016/j.avsg.2009.11.011
© 2010 Annals of Vascular Surgery Inc. Published by Elsevier Inc All rights reserved.
